Minimizing communication cost for demand response using state estimation

Demand Response (DR) is widely considered to be an integral part of future energy systems with high renewable energy penetration. Fast and reliable DR schemes require appropriate communication infrastructure. By limiting both the data sent and the measurements taken, the marginal cost of adding a load unit to the DR scheme can be reduced so that it becomes economically viable to utilize small loads such as hot-water boilers for DR. This paper presents an estimation and control topology for DR relying only on measurements of aggregated power and one-way communication. By using a state observer based on particle filtering and by switching individual loads based on comparing a drawn random number to a broadcast switching signal, accurate tracking of a reference signal can be achieved. In a case study it is shown that employing the proposed algorithms on a basic Smart Meter infrastructure enables the utility to significantly reduce overall cost by avoiding consumption of balancing energy.